Recovery of sparsest signals via ℓq-minimization

In this paper, it is proved that every s-sparse vector x∈Rn can be exactly recovered from the measurement vector z=Ax∈Rm via some ℓq-minimization with 0